Risk Mapping for Forest Pests Kurt W. Gottschalk and Andrew M. Liebhold USDA Forest Service...
-
Upload
shanna-davis -
Category
Documents
-
view
213 -
download
0
Transcript of Risk Mapping for Forest Pests Kurt W. Gottschalk and Andrew M. Liebhold USDA Forest Service...
Risk Mapping for Forest Pests
Kurt W. Gottschalk and Andrew M. Liebhold
USDA Forest Service Northeastern Research StationMorgantown, WV USA
In the Beginning…
Early 1990’s, Gypsy Moth EIS Team EIS Team wanted a national susceptibility
map or host risk mapThey contracted with Sandy and myself to
produce itWe spent about 18 months to do so
finishing in about 1994-95
In the Beginning…
Getting the data was a major task that took about a year and cost a lot of money as the FIA units charged us to create the data files for each stateFor the east it was the precursor to the east wide
database – we obtained each state independently and created a big data file
For the west, things were much worse, we could get state and private land but not national forest land
We eventually got very old, NF data from the regions and Ft. Collins
In the Beginning…
Once we got the data, the bigger problem was how to represent it because:There was no spatial data associated with the
plots other than county or ranger district for some of the western data
So we resorted to using county as our pixel size which meant we also had a problem with representing how much forest land was in a county
Most Common Preferred Tree Most Common Preferred Tree SpeciesSpecies
Common Name Genus Species Total BA(ft2/ac)
white oak Quercus alba 1425469238sweetgum Liquidambar styraciflua 1160080502quaking aspen Populus tremuloides 1008381226northern red oak Quercus rubra 961704056black oak Quercus velutina 730510718chestnut oak Quercus prinus 684442053post oak Quercus stellata 547079960water oak Quercus nigra 433745718paper birch Betula papyrifera 381347899southern red oak Quercus falcata 375025826
Northern Red Oak, Northern Red Oak, Quercus rubraQuercus rubra
00 - 0.10.1 - 0.50.5 - 2.02.0 40.0no data
ft2 / acre
Total Basal Area of Preferred SpeciesTotal Basal Area of Preferred Species
0 - 0.330.33 - 3.63.6 - 10.610.6 - 18.318.3 - 70.1no data
ft2 / acre
Second Effort….
In round two, S&PF was interested in making a national forest pest risk map
Sandy and I were approached about revising our gypsy moth effort
For this effort, we started with the FIA AVHRR forest type map
GM rate of spread was added into this effort as well
FIA forest type group map Subset for forest type groups that contain susceptible forest types (Oak-pine, Oak-hickory, Oak-gum-cypress, Elm-ash-cottonwood, and Aspen-birch):
Susceptible types
FIA Plot Data
… … excluded any counties where less than 10% of land excluded any counties where less than 10% of land area was covered by forests that have > 20% BA area was covered by forests that have > 20% BA preferred species…preferred species…
(from: Liebhold et al. J. Forestry 95: 20-24)
Third effort…
Sandy had been introduced to geostatistics, so the availability of actual GPSed plot locations allowed us to use kriging as a new approach
#
#
#
#
#
#
# #
# ###
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
# ## #
#
#
# ##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
# #
#
#
#
# #
#
#
#
# ##
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
##
# #
#
# ##
#
#
#
#
#
#
#
#
#
#
# #
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
##
##
#
#
#
#
#
##
##
#
#
#
#
#
#
# #
#
#
#
##
#
#
##
#
##
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
##
#
#
#
# ##
#
##
#
#
# #
#
#
#
#
#
#
## #
##
#
#
##
##
#
# #
#
#
# #
#
#
#
#
##
##
#
# ##
#
#
#
#
#
#
#
#
#
#
#
##
#
#
## #
#
#
# #
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
# #
##
#
#
#
#
#
#
##
#
# ##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
##
#
#
#
#
#
#
#
#
## #
#
##
#
#
#
#
#
##
#
#
##
#
##
#
#
##
#
##
#
#
#
###
#
#
#
####
##
#
#
#
#
#
##
#
#
###
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
##
#
#
##
#
# #
#
# ##
## #
#
#
#
#
#
#
##
#
###
# #
#
## #
##
##
#
#
##
#
#
#
#
##
#
# #
# ##
# #
#
#
#
#
#
####
#
#
#
##
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
# ###
#
#
#
#
#
#
##
#
# ##
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
###
#
#
#
#
#
#
##
#
##
#
#
##
##
##
#
#
##
#
#
#
#
#
#
#
##
##
#
#
#
###
#
#
#
#
#
#
###
##
#
##
#
# #
##
#
#
#
#
#
#
#
##
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
##
##
#
##
#
#
#
##
#
##
##
## ##
##
#
##
#
### ###
####
#
#
#
##
#
#
#
##
#
#
#
##
##
#
#
#
#
#
##
#
##
#
##
# #
#
##
#
#
# #
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
##
#
# #
#
#
# #
#
#
#
##
#
# #
## #
#
#
#
#
#
#
##
#
#
#
#
#
# #
#
#
#
#
##
#
#
#
#
#
#
#
#
#
# #
#
#
#
#
##
#
#
#
#
#
##
#
#
#
##
##
#
##
#
#
#
#
#
#
##
# ##
#
#
##
#
#
#
#
##
##
#
#
# #
#
#
#
#
#
# #
#
# #
#
#
##
###
# #
#
#
##
#
#
#
#
# ##
#
#
#
#
#
##
#
#
#
##
###
##
#
#
#
##
#
##
#
##
##
#
##
#
##
#
#
#
#
#
#
# # #
#
#
#
#
## #
#
#
#
# ##
##
#
#
#
##
#
###
#
##
#
#
#
#
## #
##
#
#
#
## #
#
#
#
#
#
#
#
##
#
#
#
##
#
#
#
#
#
# #
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
##
###
###
#
#
#
#
#
##
#
####
#
##
#
#
###
#
#
#
##
#
#
##
#
#
#
##
#
# #
# #
#
##
#
##
##
#
##
#
#
#
##
#
#
#
##
#
#
##
#
##
##
#
#
# #
#
## #
#
#
##
# #
#
#
#
#
#
##
##
#
#
#
#
#
#
#
###
# #
##
#
#
#
#
#
##
#
#
#
#
#
#
#
#
# #
#
#
#
#
#
##
#
#
#
#
#
#
#
#
##
##
#
#
##
##
#
##
#
##
#
##
#
#
#
#
#
#
#
#
##
#
##
#
# #
#
##
#
#
#
#
#
#
#
#
#
##
#
#
#
#
##
##
#
#
#
#
#
#
##
#
#
#
#
#
#
#
# #
#
#
#
#
#
#
#
#
#
#
#
#
## #
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
###
##
# #
#
#
#
#
# #
#
# ##
#
#
#
#
#
#
# ##
#
#
#
#
##
#
#
#
#
#
#
#
#
##
##
#
##
#
#
#
#
#
#
#
##
##
#
##
#
##
#
#
#
##
#
#
#
##
##
# #
#
#
##
##
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
##
##
# ##
#
## #
##
# #
## #
#
## #
# #
#
#
#
##
#
##
##
#
#
#
#
#
#
##
#
#
##
##
#
##
##
##
##
##
#
#
## #
#
#
#
#
##
#
#
## #
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
# #
#
#
#
##
##
#
#
#
#
##
# #
##
#
#
# # ##
#
#
#
##
##
#
#
#
# ###
##
#
##
#
##
#
#
#
#
##
# #
#
#
#
#
#
# #
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
##
##
#
#
#
#
#
##
##
#
#
#
#
#
#
#
##
#
##
##
##
#
#
#
# #
##
#
#
#
#
#
#
#
#
# #
#
#
#
#
##
#
#
##
#
#
#
#
#
#
#
#
####
#
#
#
#
#
##
#
#
#
#
#
##
##
#
#
#
# #
#
#
##
#
#
#
#
##
#
#
#
#
##
#
##
##
##
#
##
## #
#
#
##
#
#
#
#
#
#
#
##
#
#
#
#
#
# #
##
#
# ##
##
##
#
#
#
##
# #
##
#
###
#
# ##
#
#
##
#
#
#
##
# ##
##
##
#
#
#
#
#
#
##
#
#
#
#
##
# #
#
#
#
#
#
##
#
##
##
#
#
#
#
#
#
#
#
#
# #
# #
##
#
##
#
##
##
#
##
#
###
#
#
#
#
#
#
###
# #
#
#
# ##
#
#
#
#
## #
#
#
#
###
#
#
#
#
#
#
#
#
#
##
##
#
#
#
#
#
#
#
###
#
##
#
##
##
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
# #
#
#
# ##
#
#
#
#
#
##
###
####
# #
##
##
##
#
#
# #
#
##
#
#
# ##
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
#
##
#
##
#
#
## #
##
#
#
##
#
###
#
# #
#
#
# #
##
#
#
# ##
###
#
#
###
#
##
##
##
#
###
#
##
#
##
##
#
##
#
##
#
#
#
#
# #
# ##
##
#
##
#
##
# #
#
##
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
##
#
# #
#
#
# #
##
#
#
#
#
#
#
# ##
#
#
# #
#
### #
#
#
##
#
#
##
#
#
#
## #
#
#
# #
# ##
#
#
#
#
#
#
##
#
#
#
#
#
#
#
###
#
# ##
#
#
#
#
#
##
#
#
#
#
#
#
#
##
#
##
#
#
#
#
#
#
#
##
#
#
#
##
#
#
#
#
#
##
#
##
#
##
#
# #
##
#
#
#
#
# #
#
#
##
#
# ##
#
#
#
#
##
## #
##
#
# #
#
#
#
#
#
#
#
#
# #
#
#
#
#
#
#
##
# # ###
#
#
#
#
#
#
#
#
# #
#
#
#
#
#
#
##
#
#
##
#
#
#
##
#
#
#
#
##
#
#
##
#
#
#
#
#
#
##
#
#
#
##
#
#
#
#
#
#
#
##
#
#
#
##
#
# #
#
##
##
#
#
##
#
#
#
#
#
###
#
###
#
# # #
#
##
#
#
#
#
#
#
#
#
#
#
#
###
#
##
##
#
#
#
#
#
#
#
#
###
#
#
##
##
# #
##
#
# ## #
#
##
##
#
#
#
#
##
#
#
#
#
#
#
#
#
#
#
#
##
#
#
#
#
##
##
#
#
#
# ### #
# # ######
#
## #### ##
######
# ####
#
# #
### ##
### # ### # # ## #
##### # # ###
# ## ##
# ##
#
#
#
##
###
## ##
### #
### #
# # ### # #
#####
# #
##
#
#
#
###
# # ######
##### #
### ####
## ##
### # ##
#
#
#
% BA in Preferred Species
#####
0 - 10 .510 .5 - 2 7 .627 .6 - 4 7 .747 .7 - 7 0 .870 .8 - 1 00
1989 Pennsylvania FIA data % basal area:1989 Pennsylvania FIA data % basal area:Preferred by gypsy mothPreferred by gypsy moth
0-1516-2728-3940-5253-77
% BA Preferred by the Gypsy Moth % BA Preferred by the Gypsy Moth Kriged from 1989 PA FIA Plot DataKriged from 1989 PA FIA Plot Data
STDP Proposal for Risk Mapping Technology
Armed with actual plot locations and our new geostatistical tools, we got funded to develop this technology along with rate of spread into a prototype for the National Pest Risk Map
METHODS – Forest Susceptibility
The geographical distributions for each pest were mapped by interpolation of host species abundance estimated from 93,611
FIA plots located throughout the eastern U.S The ordinary kriging procedure was used to interpolate a surface
of basal area/ha for the host species of each disturbance agent We generated maps from the plot data by calculating kriged
estimates on a grid of 1- by 1-km cells Forest susceptibility maps were then adjusted for forest density The kriged host abundance maps were multiplied by the forest
density map to generate forest susceptibility maps adjusted for percent forest cover
METHODS – Spread Prediction
A spatial representation of the predicted future range expansion of each of the pests was created by estimating spread from historical records and using these estimates to predict future spread
The rate of spread was estimated as the slope of the linear model of each county’s distance from the initially infested area as a function of the time until establishment in that county
These estimated spread rates were used to generate maps representing the proportion of years of expected presence from 2002 to 2025 in 1- by 1-km grids for each agent
The proportion maps were multiplied by the adjusted forest susceptibility maps to create maps of risk for each pest through 2025
These maps depict risk as an index that includes both forest composition and the expected time the pest is expected to be present
About this time, Sudden Oak Death hit the scene
I heard a talk by David Rizzo on his tests of northern red oak and black oak
I decided to use this approach to determine the risk to the East from SOD
Our risk map went on to be used as the basis of the national SOD risk map
We then added in the NE shrub data as an additional risk factor
Kriged map of the estimated percent forest basal area for the
red and live oak groups adjusted for forest density.
Estimated Percent Forest Basal Area
Forest Health Monitoring, Evaluation Monitoring Proposal on Butternut
Butternut canker has been devastating butternut However, Mike Ostry has some putative
resistant genotypes If true resistance exists, then knowledge on
where to restore butternut is needed We mapped the occurrence of all butternut (live
and dead) and also analyzed by ecoregions and found out some interesting things
0.007530M231
0.004740M222
0.00500411
0.004610332
0.001580331
0.006150255
0.0012670234
0.01136592232
0.06140648231
0.5824321142212
0.63415626251
0.96291528M212
1.32561474M221
1.39631888221
2.0913862290222
% of Plots
w/ Butternut
Total # of
FIA Plots
# of FIA Plots with ButternutProvince
Butternut occurrence by ecoregion province.
A CART analysis of province-level proportion of plots with butternut produced
four significantly different groups.
All other Provinces
Mean = 0.09%
Provinces 222, M221, 221, M212
Mean = 1.15%
Provinces 222,M221, 221
Mean = 1.2%
Province M212Mean = 0.7%
Provinces 212,251
Mean = 0.4%
All other Provinces
Mean = 0.008%
All provincesMean = 0.37%
1.68238740M221A
1.7955910M221C
1.9691918M221B
2.0648510221F
2.0748410M212C
2.212265251B
2.31142833212K
2.41133032212F
2.78100628222J
2.8456316222I
2.9251415222H
3.43134246222M
4.3734315221B
6.3921914212E
10.901211132222L
% of Plots
w/ Butternut
Total # of
FIA Plots
# of FIA Plots with ButternutSection
Butternut occurrence by ecoregion section.
State & Private wanted an Emerald Ash Borer Risk Map
Based on our success creating the SOD risk map, an EAB risk map was requested
We did two different host layers for this map – an upland ash layer based primarily on FIA plots and a lowland ash layer based on FIA plots and other factors
Summary
Over time, we have increased our ability to create realistic estimates of species occurrences that allows us to estimate invasive pest risk
Given the host species range, we can produce a host risk map for almost any forest pest
We have used only periodic FIA data so far – the challenge ahead is to figure out how to incorporate annual FIA data into this system